Expo Data Migration Planner

Expo Data Migration Planner turns an AsyncStorage, local persistence, or saved-data schema change into a cautious migration plan before code is edited. It maps the current and target schemas, defines version detection and stepwise transformations, identifies data-loss and regression risks, covers corrupt, partial, mixed-version, and interrupted data, and checks whether the migration is idempotent.

The output includes fallback and rollback limits, a practical test matrix, and one paste-ready first task brief for Codex, Claude Code, Cursor, Windsurf, or a similar AI coding agent. It follows the existing persistence pattern when supplied, avoids unrelated refactoring and unnecessary dependencies, and does not claim that data is safe or a migration succeeded without evidence.

Within the Expo AI Coding Toolkit, this skill focuses specifically on preserving and transforming persisted data.